opencode-orchestrator 0.9.29 → 0.9.33
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +53 -30
- package/package.json +1 -1
package/README.md
CHANGED
|
@@ -27,40 +27,53 @@ Then in OpenCode:
|
|
|
27
27
|
|
|
28
28
|
## 🧠 Core Philosophy
|
|
29
29
|
|
|
30
|
-
> **Explore → Adapt → Act**
|
|
31
|
-
> *Like a human exploring unknown space —
|
|
32
|
-
|
|
33
|
-
Behaves like an astronaut exploring the unknown.
|
|
34
|
-
In unfamiliar environments: **detect without assuming**, **adapt to discoveries**, **act with confidence**.
|
|
30
|
+
> **Explore → Learn → Adapt → Act**
|
|
31
|
+
> *Like a human exploring unknown space — discover, document, adjust, execute.*
|
|
35
32
|
|
|
36
33
|
```
|
|
37
|
-
|
|
38
|
-
│
|
|
39
|
-
│
|
|
40
|
-
│
|
|
41
|
-
│
|
|
42
|
-
│
|
|
43
|
-
│
|
|
44
|
-
|
|
34
|
+
┌──────────────────────────────────────────────────────────────────────────────────────┐
|
|
35
|
+
│ │
|
|
36
|
+
│ 🔍 EXPLORE → 📝 LEARN → 🔄 ADAPT → ⚡ ACT │
|
|
37
|
+
│ │
|
|
38
|
+
│ Scan unknown Document Adjust Execute with │
|
|
39
|
+
│ territory discoveries behavior confidence │
|
|
40
|
+
│ │
|
|
41
|
+
│ • Read context • Record patterns • Match style • Build it │
|
|
42
|
+
│ • Detect stack • Note conventions • Fit the project • Test it │
|
|
43
|
+
│ • Find commands • Log findings • Apply learnings • Seal it │
|
|
44
|
+
│ │
|
|
45
|
+
└──────────────────────────────────────────────────────────────────────────────────────┘
|
|
45
46
|
```
|
|
46
47
|
|
|
47
|
-
###
|
|
48
|
+
### 🛸 The Astronaut Principle
|
|
49
|
+
|
|
50
|
+
Agents behave like astronauts landing on unknown planets:
|
|
51
|
+
|
|
52
|
+
- **Never assume gravity** — Always detect the environment first
|
|
53
|
+
- **Document everything** — Record discoveries for future missions
|
|
54
|
+
- **Adapt to atmosphere** — Adjust behavior based on learnings
|
|
55
|
+
- **Act with confidence** — Execute precisely once the terrain is mapped
|
|
56
|
+
|
|
57
|
+
This is **real-time adaptation**. Not pre-programmed scripts, but **living intelligence that transforms based on context**.
|
|
48
58
|
|
|
49
|
-
|
|
50
|
-
|
|
51
|
-
|
|
|
52
|
-
|
|
53
|
-
|
|
|
54
|
-
|
|
|
59
|
+
### 🎭 Agents & Their Adaptive Behavior
|
|
60
|
+
|
|
61
|
+
| Agent | Mission | Adaptive Behavior |
|
|
62
|
+
|:-----:|:--------|:------------------|
|
|
63
|
+
| 🎯 **Commander** | Orchestration | Discover project reality → Delegate dynamically → Loop until sealed |
|
|
64
|
+
| 📋 **Planner** | Strategy | Research actual docs → Plan parallel paths → Document findings |
|
|
65
|
+
| 🔨 **Worker** | Implementation | Observe patterns → Learn conventions → Implement with fit |
|
|
66
|
+
| ✅ **Reviewer** | Verification | Read real standards → Run actual tests → Approve with evidence |
|
|
55
67
|
|
|
56
68
|
### Guiding Principles
|
|
57
69
|
|
|
58
|
-
| Principle | Practice |
|
|
59
|
-
|
|
60
|
-
| 🔍 **Discover, Don't Assume** | Read `context.md`, detect
|
|
61
|
-
|
|
|
62
|
-
|
|
|
63
|
-
|
|
|
70
|
+
| Principle | Adaptive Practice |
|
|
71
|
+
|:----------|:------------------|
|
|
72
|
+
| 🔍 **Discover, Don't Assume** | Read `context.md`, detect stack, find actual commands |
|
|
73
|
+
| 📝 **Learn & Document** | Record patterns, conventions, and findings for reuse |
|
|
74
|
+
| 📚 **Evidence Over Memory** | Cite docs, run commands, complete only with proof |
|
|
75
|
+
| 🔄 **Transform with Context** | Behavior evolves based on documented learnings |
|
|
76
|
+
| 🤝 **Separation of Concerns** | Each agent masters one domain excellently |
|
|
64
77
|
|
|
65
78
|
---
|
|
66
79
|
|
|
@@ -106,6 +119,7 @@ In unfamiliar environments: **detect without assuming**, **adapt to discoveries*
|
|
|
106
119
|
| 🔗 **Real-Time Sync** | Shared `.opencode/` state across all agents |
|
|
107
120
|
| 🛡️ **Auto Verification** | E2E tests, import checks, integration validation |
|
|
108
121
|
| 🩹 **Self-Healing** | Auto-recovery with 3 retries per session |
|
|
122
|
+
| 🧬 **Adaptive Intelligence** | Agents evolve behavior based on discoveries |
|
|
109
123
|
|
|
110
124
|
### Self-Healing Details
|
|
111
125
|
|
|
@@ -141,11 +155,20 @@ In unfamiliar environments: **detect without assuming**, **adapt to discoveries*
|
|
|
141
155
|
|
|
142
156
|
---
|
|
143
157
|
|
|
144
|
-
##
|
|
158
|
+
## 🎹 Developer's Words
|
|
159
|
+
|
|
160
|
+
<div align="center">
|
|
161
|
+
<img src="assets/image.png" alt="Piano and Code" width="600" />
|
|
162
|
+
</div>
|
|
163
|
+
|
|
164
|
+
<br />
|
|
165
|
+
|
|
166
|
+
> *I believe playing the piano is also a form of orchestration.*
|
|
167
|
+
>
|
|
168
|
+
> *The harmony of polyphony — multiple voices — and homophony — a single melodic line.*
|
|
169
|
+
>
|
|
170
|
+
> *Each voice sings its most beautiful song from its own place, yet when combined, they create one grand, beautiful melody. I believe this structure is no different from AI agents.*
|
|
145
171
|
|
|
146
|
-
```bash
|
|
147
|
-
npm uninstall -g opencode-orchestrator
|
|
148
|
-
```
|
|
149
172
|
|
|
150
173
|
---
|
|
151
174
|
|
package/package.json
CHANGED
|
@@ -2,7 +2,7 @@
|
|
|
2
2
|
"name": "opencode-orchestrator",
|
|
3
3
|
"displayName": "OpenCode Orchestrator",
|
|
4
4
|
"description": "Distributed Cognitive Architecture for OpenCode. Turns simple prompts into specialized multi-agent workflows (Planner, Coder, Reviewer).",
|
|
5
|
-
"version": "0.9.
|
|
5
|
+
"version": "0.9.33",
|
|
6
6
|
"author": "agnusdei1207",
|
|
7
7
|
"license": "MIT",
|
|
8
8
|
"repository": {
|